Hi Piotr, Hi Becket, Hi everybody, we, Dawid and I, discussed the various suggestions/options and we would be okay either way because we find neither solution is perfect just because of the already present complexity.
Option c) Adding methods to the interfaces of SourceReader and SplitReader Option a) Adding decorative interfaces to be used by SourceReader and SplitReader As of the current status (v. 12) of the FLIP [1], it is based on Option c) which we find acceptable because the complexity added is only a single method. Nevertheless, Piotr has valid concerns about Option c) which I see as follows: (1) An interface with default NOOP implementation makes the implementation optional. And in my opinion, a default implementation is and will remain a way of making implementation optional because even in future a developer can decide to implement the "old flavor" without support for pausable splits. (2) It may not be too critical but I also find it suboptimal that with a NOOP default implementation there is no way to check at runtime if SourceReader or SplitReader actually support pausing. (To do so, one would need a supportsX method which makes it again more complicated.) However, we haven't changed it because Option a) is also not optimal or straight-forward: (1) We need to add two distinct yet similar decorative interfaces since, as mentioned, the signatures of the methods are different. For example, we would need decorative interfaces like `SplitReaderWithPausableSplits` and `SourceReaderWithPausableSplits`. (2) As a consequence, we would need to somehow document how/where to implement both interfaces and how this relates to each other. This we could solve by adding a note in the interface of SourceReader and SplitReader and reference to the decorative interfaces but it still increases complexity too. In summary, we see both as acceptable and preferred over other options. The question is if we can find a solution or compromise that is acceptable for everybody to reach consensus. Please let us know what you think because we would be happy if we can conclude the discussion to avoid dropping the initiative on this FLIP. Regards, Sebastian [1] https://cwiki.apache.org/confluence/pages/viewpage.action?pageId=199540438 (v. 12) On Thu, May 5, 2022 at 10:13 AM Piotr Nowojski <pnowoj...@apache.org> wrote: > Hi Guowei, > > as Dawid wrote a couple of messages back: > > > This is covered in the previous FLIP[1] which has been already > implemented in 1.15. In short, it must be enabled with the watermark > strategy which also configures drift and update interval > > So by default watermark alignment is disabled, regardless if a source > supports it or not. > > Best, > Piotrek > > czw., 5 maj 2022 o 09:56 Guowei Ma <guowei....@gmail.com> napisał(a): > >> Hi, >> >> We know that in the case of Bounded input Flink supports the Batch >> execution mode. Currently in Batch execution mode, flink is executed on a >> stage-by-stage basis. In this way, perhaps watermark alignment might not >> gain much. >> >> So my question is: Is watermark alignment the default behavior(for >> implemented source only)? If so, have you considered evaluating the impact >> of this behavior on the Batch execution mode? Or thinks it is not >> necessary. >> >> Correct me if I miss something. >> >> Best, >> Guowei >> >> >> On Thu, May 5, 2022 at 1:01 PM Piotr Nowojski <piotr.nowoj...@gmail.com> >> wrote: >> >> > Hi Becket and Dawid, >> > >> > > I feel that no matter which option we choose this can not be solved >> > entirely in either of the options, because of the point above and >> because >> > the signature of SplitReader#pauseOrResumeSplits and >> > SourceReader#pauseOrResumeSplits are slightly different (one identifies >> > splits with splitId the other one passes the splits directly). >> > >> > Yes, that's a good point in this case and for features that need to be >> > implemented in more than one place. >> > >> > > Is there any reason for pausing reading from a split an optional >> feature, >> > > other than that this was not included in the original interface? >> > >> > An additional argument in favor of making it optional is to simplify >> source >> > implementation. But on its own I'm not sure if that would be enough to >> > justify making this feature optional. Maybe. >> > >> > > I think it would be way simpler and clearer to just let end users and >> > Flink >> > > assume all the connectors will implement this feature. >> > >> > As I wrote above that would be an interesting choice to make (ease of >> > implementation for new users, vs system consistency). Regardless of >> that, >> > yes, for me the main argument is the API backward compatibility. But >> let's >> > clear a couple of points: >> > - The current proposal adding methods to the base interface with default >> > implementations is an OPTIONAL feature. Same as the decorative version >> > would be. >> > - Decorative version could implement "throw >> UnsupportedOperationException" >> > if user enabled watermark alignment just as well and I agree that's a >> > better option compared to logging a warning. >> > >> > Best, >> > Piotrek >> > >> > >> > śr., 4 maj 2022 o 15:40 Becket Qin <becket....@gmail.com> napisał(a): >> > >> > > Thanks for the reply and patient discussion, Piotr and Dawid. >> > > >> > > Is there any reason for pausing reading from a split an optional >> feature, >> > > other than that this was not included in the original interface? >> > > >> > > To be honest I am really worried about the complexity of the user >> story >> > > here. Optional features like this have a high overhead. Imagine this >> > > feature is optional, now a user enabled watermark alignment and >> defined a >> > > few watermark groups. Would it work? Hmm, that depends on whether the >> > > involved Source has implmemented this feature. If the Sources are well >> > > documented, good luck. Otherwise end users may have to look into the >> code >> > > of the Source to see whether the feature is supported. Which is >> something >> > > they shouldn't have to do. >> > > >> > > I think it would be way simpler and clearer to just let end users and >> > Flink >> > > assume all the connectors will implement this feature. After all the >> > > watermark group is not optinoal to the end users. If in some rare >> cases, >> > > the feature cannot be supported, a clear UnsupportedOperationException >> > will >> > > be thrown to tell users to explicitly remove this Source from the >> > watermark >> > > group. I don't think we should have a warning message here, as they >> tend >> > to >> > > be ignored in many cases. If we do this, we don't even need the >> > supportXXX >> > > method in the Source for this feature. In fact this is exactly how >> many >> > > interfaces works today. For example, SplitEnumerator#addSplitsBack() >> is >> > not >> > > supported by Pravega source because it does not support partial >> failover. >> > > In that case, it simply throws an exception to trigger a global >> recovery. >> > > >> > > The reason we add a default implementation in this case would just for >> > the >> > > sake of backwards compatibility so the old source can still compile. >> > Sure, >> > > in short term, this feature might not be supported by many existing >> > > sources. That is OK, and it is quite visible to the source developers >> > that >> > > they did not override the default impl which throws an >> > > UnsupportedOperationException. >> > > >> > > @Dawid, >> > > >> > > the Java doc of the SupportXXX() method in the Source would be the >> single >> > > >> source of truth regarding how to implement this feature. >> > > > >> > > > >> > > >> > > I also don't find it entirely true. Half of the classes are >> theoretically >> > > > optional and are utility classes from the point of view how the >> > > interfaces >> > > > are organized. Theoretically users do not need to use any of >> > > > SourceReaderBase & SplitReader. Would be weird to list their >> methods in >> > > the >> > > > Source interface. >> > > >> > > I think the ultimate goal of java docs is to guide users to implement >> the >> > > Source. If SourceReaderBase is the preferred way to implement a >> > > SourceReader, it seems worth mentioning that. Even the Java language >> > > documentation interfaces lists the konwn implementations [1] so people >> > can >> > > leverage them. But for this particular case, if we make the feature >> > > non-optional, we don't even need the supportXXX() method for now. >> > > >> > > Thanks, >> > > >> > > Jiangjie (Becket) Qin >> > > >> > > >> > > >> > > On Wed, May 4, 2022 at 4:37 PM Dawid Wysakowicz < >> dwysakow...@apache.org> >> > > wrote: >> > > >> > > > Hey Piotr and Becket, >> > > > >> > > > First of all, let me say I am happy with whichever option is agreed >> in >> > > the >> > > > discussion. >> > > > >> > > > I wanted to clarify a few points from the discussion though: >> > > > >> > > > @Becket: >> > > > >> > > > The main argument for adding the methods to the SourceReader is that >> > > these >> > > > methods are effectively NON-OPTIONAL to SourceReader impl, i.e. >> > starting >> > > > from this FLIP, all the SourceReaders impl are expected to support >> this >> > > > method, although some old implementations may not have implemented >> this >> > > > feature. I think we should distinguish the new features from the >> > optional >> > > > features. While the public decorative interface is a solution to the >> > > > optional features. We should not use it for the features that are >> > > > non-optional. >> > > > >> > > > I don't think that this feature is NON-OPTIONAL. Even though >> > preferred, I >> > > > still think it can be simply optional. >> > > > >> > > > the Java doc of the SupportXXX() method in the Source would be the >> > single >> > > > source of truth regarding how to implement this feature. >> > > > >> > > > I also don't find it entirely true. Half of the classes are >> > theoretically >> > > > optional and are utility classes from the point of view how the >> > > interfaces >> > > > are organized. Theoretically users do not need to use any of >> > > > SourceReaderBase & SplitReader. Would be weird to list their >> methods in >> > > the >> > > > Source interface. >> > > > >> > > > @Piotr >> > > > >> > > > If we have all of the methods with default implementation in the >> base >> > > > interface, the API doesn't give any clue to the user which set of >> > methods >> > > > are required to be implemented at the same time. >> > > > >> > > > I feel that no matter which option we choose this can not be solved >> > > > entirely in either of the options, because of the point above and >> > because >> > > > the signature of SplitReader#pauseOrResumeSplits and >> > > > SourceReader#pauseOrResumeSplits are slightly different (one >> identifies >> > > > splits with splitId the other one passes the splits directly). >> > > > >> > > > Best, >> > > > >> > > > Dawid >> > > > On 03/05/2022 14:30, Becket Qin wrote: >> > > > >> > > > Hi Piotr, >> > > > >> > > > Thanks for the comment. >> > > > >> > > > Just to clarify, I am not against the decorative interfaces, but I >> do >> > > > think we should use them with caution. The main argument for adding >> the >> > > > methods to the SourceReader is that these methods are >> > > > effectively NON-OPTIONAL to SourceReader impl, i.e. starting from >> this >> > > > FLIP, all the SourceReaders impl are expected to support this >> > > > method, although some old implementations may not have implemented >> this >> > > > feature. I think we should distinguish the new features from the >> > optional >> > > > features. While the public decorative interface is a solution to the >> > > > optional features. We should not use it for the features that are >> > > > non-optional. >> > > > >> > > > That said, this feature is optional for SplitReaders. Arguably we >> can >> > > have >> > > > a decorative interface for that, but for simplicity and symmetry of >> the >> > > > interface, personally I prefer just adding a new method. >> > > > >> > > > Regarding the advantages you mentioned about the decorative >> interfaces, >> > > > they would make sense if: >> > > > 1. The feature is optional. >> > > > 2. There is only one decorative interface involved for a feature. >> > > > Otherwise the argument that all the methods are grouped together >> will >> > not >> > > > stand. >> > > > >> > > > Compared with that, I think the current solution works fine in all >> > cases, >> > > > i.e. "having supportXXX() method in Source, and default methods / >> > > > decorative interfaces in base interfaces.". >> > > > >> > > > The advantages are: >> > > >> - clean and easy to implement base interface >> > > > >> > > > In the current approach, the Java doc of the SupportXXX() method in >> the >> > > > Source would be the single source of truth regarding how to >> implement >> > > this >> > > > feature. It lists the method that has to be implemented to support >> this >> > > > feature, regardless of how many classes / interfaces are involved. >> > > > >> > > > When implementing the base interface, users do not need to >> implement a >> > > > method with default implementation. If they are curious what the >> method >> > > is >> > > > for, the java doc of that method simply points users to the >> > SupportXXX() >> > > > method in the Source. It seems not adding work to the users compared >> > with >> > > > decorative interfaces, but gives much better discoverability. >> > > > >> > > > - all of the methods from a single feature are grouped in a single >> > > >> decorator interface, together with their dedicated java doc. It's >> also >> > > >> easier to google search for help using the decorator name >> > > > >> > > > - if an optional feature requires two methods to be implemented at >> > once, >> > > >> decorator can guarantee that >> > > > >> > > > These two points are not true when multiple components and classes >> are >> > > > involved collaboratively to provide a feature. In our case, we have >> > both >> > > > SourceReader and SplitReader involved. And there might be other >> > > interfaces >> > > > on the JM side involved for some future features. So the relevant >> > methods >> > > > can actually be scattered over the places. That said, we may still >> use >> > > > decorative interfaces for each component, if the feature is >> optional, >> > > given >> > > > there is a single source of truth for the feature. >> > > > >> > > > Here I would strongly lean towards making life easier for new users, >> > > >> lowering the entry barrier, at the (imo) slight expense for the >> power >> > > >> users. >> > > > >> > > > I actually think the current approach is simpler, more extensible >> and >> > > more >> > > > general for all the users. Can you articulate a bit more on which >> part >> > > you >> > > > think makes users harder to understand? >> > > > >> > > > >> > > > There is another benefit of the decorative interfaces which is not >> > > > mentioned, but might be worth considering here. Usually the >> decorative >> > > > interfaces give slightly better backwards compatibility than the new >> > > > default method in the interfaces. That is when users are using a jar >> > that >> > > > was compiled with an older version of Flink which does not have the >> > > default >> > > > method in the interfaces in question. A decorative interface may >> still >> > > > provide backwards compatibility in that case, while default method >> impl >> > > > cannot. >> > > > >> > > > I think in Flink we in general do not guarantee custom components >> > > compiled >> > > > with an older version can run with a newer version of Flink. A >> > recompile >> > > > with a newer version would be required. That said, if we do care >> about >> > > > this, we can just change the "supportXXX()" method in the Source >> > > interface >> > > > to use decorative interfaces, and leave the other parts unchanged. >> > > > >> > > > Thanks, >> > > > >> > > > Jiangjie (Becket) Qin >> > > > >> > > > >> > > > >> > > > >> > > > On Tue, May 3, 2022 at 6:25 PM Piotr Nowojski <pnowoj...@apache.org >> > >> > > > wrote: >> > > > >> > > >> Hi, >> > > >> >> > > >> Sorry for chipping in so late, but I was OoO for the last two >> weeks. >> > > >> Regarding the interfaces, I would be actually against adding those >> > > methods >> > > >> to the base interfaces for the reasons mentioned above. Clogging >> the >> > > base >> > > >> interface for new users with tons of methods that they do not >> need, do >> > > not >> > > >> understand and do not know what to do with them. Moreover, such >> > > decorative >> > > >> interfaces are solving a problem if a feature requires two or more >> > > methods >> > > >> to be implemented at the same time. If we have all of the methods >> with >> > > >> default implementation in the base interface, the API doesn't give >> any >> > > >> clue >> > > >> to the user which set of methods are required to be implemented at >> the >> > > >> same >> > > >> time. >> > > >> >> > > >> > a) I feel the biggest drawback of decorative interfaces is which >> > > >> interface >> > > >> > they can decorate and which combinations of multiple decorative >> > > >> interfaces >> > > >> > are valid. (...) >> > > >> > In the future, if there is a new feature added >> > > >> > (e.g. sorted or pre-partitioned data aware), are we going to >> create >> > > >> another >> > > >> > interface of SplitReader such as SortedSplitReader or >> > > >> PrePartitionedAware? >> > > >> > Can they be combined? So I think the additional decorative >> interface >> > > >> like >> > > >> > withSplitsAlignment actually increases the understanding cost of >> > users >> > > >> > because they have to know what decorative interfaces are there, >> > which >> > > >> > interface they can decorate and which combinations of the >> decorative >> > > >> > interfaces are valid and which are not. Ideally we want to avoid >> > that. >> > > >> >> > > >> I'm not sure if I understand how embedding default methods in the >> base >> > > >> interface is solving the problem: what can be combined or not? If >> > there >> > > >> are >> > > >> two conflicting features, having decorative interfaces that can >> not be >> > > >> mixed together actually makes much more sense to me rather than >> having >> > > >> them >> > > >> all in one base class. How would you allow users to implement only >> one >> > > of >> > > >> those two features? >> > > >> >> > > >> To reiterate on the issue. Yes, there are drawbacks: >> > > >> - how can a user discover what decorators are there? >> > > >> - how can a user know where the decorator can be applied to? >> > > >> >> > > >> However those are drawbacks for more power users, that can be >> > mitigated >> > > by >> > > >> the documentation. For example listing all of the decorators with >> > > >> detailed explanation both in the docs and in the java docs. More >> > > >> experienced users will be able to deal with those issues easier, as >> > they >> > > >> will already have some basic understanding of Flink. Also if user >> has >> > a >> > > >> problem that he wants to solve, he will google search a potential >> > > solution >> > > >> to his problem anyway, and while doing that he is very likely to >> > > discover >> > > >> the decorator that he needs anyway in the docs. >> > > >> >> > > >> The advantages are: >> > > >> - clean and easy to implement base interface >> > > >> - all of the methods from a single feature are grouped in a single >> > > >> decorator interface, together with their dedicated java doc. It's >> also >> > > >> easier to google search for help using the decorator name >> > > >> - if an optional feature requires two methods to be implemented at >> > once, >> > > >> decorator can guarantee that >> > > >> >> > > >> Here I would strongly lean towards making life easier for new >> users, >> > > >> lowering the entry barrier, at the (imo) slight expense for the >> power >> > > >> users. >> > > >> >> > > >> Best, >> > > >> Piotrek >> > > >> >> > > >> >> > > >> wt., 26 kwi 2022 o 15:32 Becket Qin <becket....@gmail.com> >> > napisał(a): >> > > >> >> > > >> > Thanks for the reply Sebastian and Dawid, >> > > >> > >> > > >> > I think Sebastion has a good summary. This is a really helpful >> > > >> discussion. >> > > >> > >> > > >> > Thinking a bit more, I feel that it might still be better to add >> the >> > > >> > supportsXXX() method in the Source rather than SourceReader. >> > > >> > >> > > >> > Generally speaking, what we are trying to do here is to let the >> > Flink >> > > >> > framework know what the Source is capable of. In this FLIP, it >> > happens >> > > >> to >> > > >> > be the capability that only involves SourceReader. But in the >> > future, >> > > >> it is >> > > >> > possible that another functionality involves both the >> > SplitEnumerator >> > > >> and >> > > >> > SourceReader. In that case, following the current approach, we >> > should >> > > >> put >> > > >> > the "supportsXXX()" method in both SplitEnumerator and >> SourceReader. >> > > >> > Because if we only put this in the SourceReader, then the JM >> would >> > > have >> > > >> to >> > > >> > create a SourceReader in order to know whether this feature is >> > > >> supported, >> > > >> > which is a little ugly. But if we put the "supportsXXX()" method >> in >> > > the >> > > >> > Source, we will break the "symmetric" design because this FLIP >> > chose a >> > > >> > different way. >> > > >> > >> > > >> > This is also why I think supportsXXX() method seems a good thing >> to >> > > >> have, >> > > >> > because when there are a few interfaces / methods that are >> expected >> > to >> > > >> be >> > > >> > implemented at the same time in order to deliver a feature, it is >> > > always >> > > >> > good to have a single source of truth to tell the framework what >> to >> > > do, >> > > >> so >> > > >> > the framework can do consistent things in different parts. >> > > >> > >> > > >> > @Sebastian Mattheis <sebast...@ververica.com> >> > > >> > >> > > >> > Regarding interface flavor b), i.e. AlignedSourceReader + >> > > >> > AlignedSplitReader, what I feel awkward about is that we are >> > > essentially >> > > >> > expecting almost all the SourceReader implementations to extend >> > > >> > SourceReaderBase, which effectively makes the SourceReader >> interface >> > > >> > without the pausing support useless. So this indicates that >> public >> > > >> > decorative interfaces (or sub-interfaces for the same purpose) >> only >> > > >> > make sense if the original interface is also expected to be used. >> > > >> > Otherwise, it seems makes more sense to add the method to the >> > original >> > > >> > interface itself. >> > > >> > >> > > >> > Cheers, >> > > >> > >> > > >> > Jiangjie (Becket) Qin >> > > >> > >> > > >> > >> > > >> > >> > > >> > >> > > >> > On Tue, Apr 26, 2022 at 6:05 PM Dawid Wysakowicz < >> > > >> dwysakow...@apache.org> >> > > >> > wrote: >> > > >> > >> > > >> > > Thanks @Sebastian for the nice summary. >> > > >> > > >> > > >> > > I think most of your points aligned with the suggestions I >> made to >> > > the >> > > >> > > FLIP, while you were writing your reply (I believe we hit enter >> > > >> nearly at >> > > >> > > the same time ;) ) >> > > >> > > >> > > >> > > Two points after we synced offline >> > > >> > > >> > > >> > > 1. I changed also the supportsWatermarksSplitAlignment to >> > > >> > > supportsPausingSplits to express the general capability of >> > pausing. >> > > >> > > >> > > >> > > 2. As for if we should PausingSourceReader/PausingSplitReader >> > > (option >> > > >> b) >> > > >> > > or if we should just add the methods (option c), I suggest to >> > simply >> > > >> add >> > > >> > > the two methods as I felt this is much preferred approach >> Becket, >> > > >> which >> > > >> > > others do not object. Unless there is an opposition let's go >> with >> > > this >> > > >> > > option c. >> > > >> > > >> > > >> > > Best, >> > > >> > > >> > > >> > > Dawid >> > > >> > > On 26/04/2022 10:06, Sebastian Mattheis wrote: >> > > >> > > >> > > >> > > Hi folks, >> > > >> > > >> > > >> > > Sorry for being a bit silent. Many thanks for all the input and >> > > >> > > suggestions. As I'm a bit new, I needed some time to catch up >> and >> > > >> > structure >> > > >> > > (for myself) the discussion and I wanted to find a way to >> > structure >> > > >> the >> > > >> > > conclusions. (Also because I had the feeling that some concerns >> > got >> > > >> lost >> > > >> > in >> > > >> > > the discussion.) This is my attempt and please correct me if >> > > >> something is >> > > >> > > wrong or misunderstood. I tried to collect and assemble the >> > > opinions, >> > > >> > > suggestions, and conclusions (to the best of my knowledge): >> > > >> > > >> > > >> > > # Top A: Should split alignment (pause/resume behavior) be a >> > general >> > > >> > > capability? >> > > >> > > >> > > >> > > I personally don't see any reason no to have it a general >> > capability >> > > >> > > because for the alignSplit method it is actually independent of >> > the >> > > >> > > watermarks. If we agree here to have it a general capability, >> we >> > > >> should >> > > >> > > also agree on the right wording. Does >> "alignSplits(splitsToResume, >> > > >> > > splitsToPause)" refer to what is then actually meant? (I see >> it as >> > > >> okay. >> > > >> > I >> > > >> > > don't have any better idea whilst Arvid suggested >> > > >> "pauseOrResumeSplits".) >> > > >> > > >> > > >> > > # Top B: Should it be possible do enable/disable split >> alignment? >> > > >> > > >> > > >> > > I would personally not disable the split alignment on the >> source >> > > >> reader >> > > >> > > side because if split alignment is used for some other use case >> > (see >> > > >> A) >> > > >> > it >> > > >> > > could have nasty side effects on other/future use cases. >> Instead, >> > I >> > > >> would >> > > >> > > disable "watermark split alignment" where I think it should >> > disable >> > > >> the >> > > >> > > watermark-dependent trigger for split alignment. >> > > >> > > >> > > >> > > # Top C: Should we add a supportsX method? >> > > >> > > >> > > >> > > I find it difficult to define the scope of a supportsX method >> > w.r.t. >> > > >> to >> > > >> > > the following questions: a) Where is it used? and b) What is >> the >> > > >> expected >> > > >> > > output? To b), it's not straight-forward to provide a >> meaningful >> > > >> output, >> > > >> > > e.g., if SourceReader supports split alignment but SplitReader >> > not. >> > > >> This >> > > >> > is >> > > >> > > because with the current implementation, we can determine >> whether >> > > >> split >> > > >> > > alignment is fully supported only during runtime and >> specifically >> > > >> > actually >> > > >> > > only when calling alignSplits down the call hierarchy up to the >> > > actual >> > > >> > > SplitReaders. >> > > >> > > >> > > >> > > Therefore, I would suggest to either raise an error or warning >> if >> > > the >> > > >> > > alignment is called but not supported at some point. I know we >> > > should >> > > >> > > carefully think about when this could be the case because we >> don't >> > > >> want >> > > >> > to >> > > >> > > flood anybody with such warnings. However, warnings could be an >> > > >> indicator >> > > >> > > for the user that for watermark split alignment use case split >> > > >> reading is >> > > >> > > imbalanced with the conclusion to either disable the trigger >> for >> > > >> > watermark >> > > >> > > split alignment (see Top B) or to use/implement a source and >> > reader >> > > >> that >> > > >> > > fully supports split alignment. >> > > >> > > >> > > >> > > # Top D: How to design interfaces? >> > > >> > > >> > > >> > > Thanks for structuring the discussion with the the various >> > > >> possibilities >> > > >> > > (a-d). From the discussion and emails, I would like to >> summarize >> > the >> > > >> > > following requirements: >> > > >> > > - Interfaces should be consistent ("symmetric"), i.e., similar >> > > >> semantics >> > > >> > > should have similar interfaces with similar usage. >> > > >> > > - Make explicit which implementations implement >> interfaces/support >> > > >> > > behavior. >> > > >> > > - Make clear what are default implementations and how to >> implement >> > > >> > > interfaces with desired behavior. >> > > >> > > >> > > >> > > This is a simplified view of the relations between relevant >> > classes >> > > of >> > > >> > the >> > > >> > > PoC implementation: >> > > >> > > >> > > >> > > SourceReader (Public) <|-- SourceReaderBase (Internal) <|-- .. >> > <|-- >> > > >> > > MySourceReader >> > > >> > > >> > > >> > > MySourceReader <>-- SplitFetcherManager (Internal) <>-- >> > SplitFetcher >> > > >> > > (Internal) <>-- SplitReader (Public) <|-- MySplitReader >> > > >> > > >> > > >> > > (A <|-- B: B inherits from A; A <>-- B: A "has a" B) >> > > >> > > >> > > >> > > Note that SourceReaderBase and SplitFetcherManager implement >> most >> > of >> > > >> the >> > > >> > > "logic" for split alignment just because we wanted to implement >> > > split >> > > >> > > alignment and wanted it to be available as kind of a default. >> As a >> > > >> > > consequence, we have a "default implementation" for >> SourceReader >> > > that >> > > >> > > implements the actual logic for split alignment. For that >> reason, >> > I >> > > >> find >> > > >> > it >> > > >> > > very confusing to have a NOOP default implementation in the >> > > interface >> > > >> for >> > > >> > > the SourceReader. As a consequence, interface strategy c) is >> > > difficult >> > > >> > > because this would require NOOP default implementations in the >> > > public >> > > >> > > interfaces of SourceReader and SplitReader. This is the same >> for >> > > >> strategy >> > > >> > > d) because it would require NOOP default implementation in the >> > > >> > > SourceReader. Further, as Dawid described method signatures of >> > > >> alignSplit >> > > >> > > for SourceReader and SplitReader differ and it would be >> extremely >> > > >> > difficult >> > > >> > > to make the signatures the same (with even potential >> performance >> > > >> impact >> > > >> > > because of additional loop-ups of split ids). Therefore, >> having a >> > > >> > symmetric >> > > >> > > decorative interface as of strategy a) is actually not possible >> > and >> > > >> > having >> > > >> > > two decorative interfaces with different method signatures is >> > > >> confusing. >> > > >> > My >> > > >> > > conclusion is that we are best with strategy b) which means to >> > have >> > > >> > > specializing sub-interfaces that inherit from the parent >> > interface: >> > > >> > > SourceReader <|-- AlignedSourceReader, SplitReader <|-- >> > > >> > AlignedSplitReader >> > > >> > > With this option, I'm not 100% sure what the implications are >> and >> > if >> > > >> this >> > > >> > > could get nasty. I would suggest that Dawid and I just try to >> > > >> implement >> > > >> > and >> > > >> > > see if we like it. :) >> > > >> > > >> > > >> > > # Summary >> > > >> > > >> > > >> > > In conclusion, please let me know your perspectives. Please >> > correct >> > > >> me, >> > > >> > if >> > > >> > > something is wrong or if I misunderstood something. My >> perspective >> > > >> would >> > > >> > be: >> > > >> > > >> > > >> > > Top A: Yes >> > > >> > > Top B: Yes (but disable watermark trigger for split alignment) >> > > >> > > Top C: No >> > > >> > > Top D: b) >> > > >> > > >> > > >> > > Best, >> > > >> > > Sebastian >> > > >> > > >> > > >> > > On Tue, Apr 26, 2022 at 9:55 AM Dawid Wysakowicz < >> > > >> dwysakow...@apache.org >> > > >> > > >> > > >> > > wrote: >> > > >> > > >> > > >> > >> @Arvid: >> > > >> > >> >> > > >> > >> While I also like Becket's capability approach, I fear that it >> > > >> doesn't >> > > >> > work >> > > >> > >> for this particular use case: Sources can always be aligned >> > > >> cross-task >> > > >> > and >> > > >> > >> this is just about intra-task alignment. So it's plausible to >> put >> > > >> > sources >> > > >> > >> into an alignment group even though they do not use any of the >> > > >> presented >> > > >> > >> API of FLIP-217. They should just issue a warning, if they >> handle >> > > >> > multiple >> > > >> > >> splits (see motivation section). >> > > >> > >> >> > > >> > >> Yes, but the "supportXXX" method would be for telling if it >> > > supports >> > > >> > that >> > > >> > >> intra-task alignment. Cross-task alignment would always be >> > > supported. >> > > >> > >> >> > > >> > >> I updated interfaces to what I believe to be closest to a >> > consensus >> > > >> > >> between all participants. Do you mind taking a look? >> > > >> > >> >> > > >> > >> @Sebastian Do you mind addressing the nits? >> > > >> > >> >> > > >> > >> Best, >> > > >> > >> >> > > >> > >> Dawid >> > > >> > >> >> > > >> > >> On 25/04/2022 13:39, Arvid Heise wrote: >> > > >> > >> >> > > >> > >> Thanks for pushing this effort. >> > > >> > >> >> > > >> > >> I'd actually be in favor of 1b). I fully agree that decorator >> > > >> interfaces >> > > >> > >> should be avoided but I'm also not a big fan of overloading >> the >> > > base >> > > >> > >> interfaces (they are hard to implement as is). The usual >> feedback >> > > to >> > > >> > >> Source-related interfaces are always that they are >> overwhelming >> > and >> > > >> too >> > > >> > >> hard to implement. However, I'd also not oppose 1c) as >> scattered >> > > >> > interfaces >> > > >> > >> also have drawbacks. I'd just dislike 1a) and 1d). >> > > >> > >> While I also like Becket's capability approach, I fear that it >> > > >> doesn't >> > > >> > work >> > > >> > >> for this particular use case: Sources can always be aligned >> > > >> cross-task >> > > >> > and >> > > >> > >> this is just about intra-task alignment. So it's plausible to >> put >> > > >> > sources >> > > >> > >> into an alignment group even though they do not use any of the >> > > >> presented >> > > >> > >> API of FLIP-217. They should just issue a warning, if they >> handle >> > > >> > multiple >> > > >> > >> splits (see motivation section). >> > > >> > >> >> > > >> > >> I think renaming alignSplits to facilitate future use cases >> makes >> > > >> sense >> > > >> > but >> > > >> > >> then all interfaces (if 1c) is chosen) should be adjusted >> > > >> accordingly. >> > > >> > >> AlignedSourceReader could be PausingSourceReader and I'd go >> for >> > > >> > >> pauseOrResumeSplits (Becket's proposal afaik). We could also >> > split >> > > it >> > > >> > into >> > > >> > >> pauseSplit and resumeSplit. While pauseOrResumeSplits may >> allow >> > > >> Sources >> > > >> > to >> > > >> > >> just use 1 instead of 2 library calls (as written in the >> > Javadoc), >> > > >> both >> > > >> > >> Kafka and Pulsar can't use it and I'm not sure if there is a >> > system >> > > >> that >> > > >> > >> can. >> > > >> > >> >> > > >> > >> Some nit for the FLIP: >> > > >> > >> - Please replace "stop" with "pause". >> > > >> > >> - Not sure if it's worth it in the capability section: Sources >> > that >> > > >> > adopt >> > > >> > >> this interface cannot be used in earlier versions. So it feels >> > like >> > > >> we >> > > >> > are >> > > >> > >> only forward compatible (old sources can be used after the >> > change); >> > > >> but >> > > >> > I >> > > >> > >> guess this holds for any API addition. >> > > >> > >> - You might want to add what happens when all splits are >> paused. >> > > >> > >> - You may want to describe how the 3 flavors of >> SourceReaderBase >> > > >> > interact >> > > >> > >> with the interface. >> > > >> > >> - I'm not sure if it makes sense to include Kafka and Pulsar >> in >> > the >> > > >> > FLIP. >> > > >> > >> For me, this is rather immediate follow-up work. (could be in >> the >> > > >> same >> > > >> > >> umbrella ticket) >> > > >> > >> >> > > >> > >> Best, >> > > >> > >> >> > > >> > >> Arvid >> > > >> > >> >> > > >> > >> On Mon, Apr 25, 2022 at 12:52 PM Dawid Wysakowicz < >> > > >> > dwysakow...@apache.org> <dwysakow...@apache.org> >> > > >> > >> wrote: >> > > >> > >> >> > > >> > >> >> > > >> > >> a) "MySourceReader implements SourceReader, >> WithSplitsAlignment", >> > > >> along >> > > >> > >> with "MySplitReader implements SplitReader, >> WithSplitsAlignment", >> > > or >> > > >> > >> b) "MySourceReader implements AlignedSourceReader" and >> > > "MySplitReader >> > > >> > >> implements AlignedSplitReader", or >> > > >> > >> c) "MySourceReader implements SourceReader" and "MySplitReader >> > > >> > implements >> > > >> > >> SplitReader". >> > > >> > >> >> > > >> > >> I think the latest proposal according to Dawid would be: >> > > >> > >> d) "MySourceReader implements SourceReader" and "MySplitReader >> > > >> > implements >> > > >> > >> AlignedSplitReader". >> > > >> > >> I am fine with this API, although personally speaking I think >> it >> > is >> > > >> > simpler >> > > >> > >> to just add a new method to the split reader with default >> impl. >> > > >> > >> >> > > >> > >> >> > > >> > >> I think that is a good idea to have it aligned as much as >> > possible. >> > > >> I'd >> > > >> > be >> > > >> > >> +1 for your option c). We can merge AlignedSplitReader with >> > > >> > SplitReader. We >> > > >> > >> will update the FLIP shortly. >> > > >> > >> >> > > >> > >> Best, >> > > >> > >> >> > > >> > >> Dawid >> > > >> > >> >> > > >> > >> On 25/04/2022 12:43, Becket Qin wrote: >> > > >> > >> >> > > >> > >> Thanks for the comment, Jark. >> > > >> > >> >> > > >> > >> 3. Interface/Method Name. >> > > >> > >> >> > > >> > >> Can the interface be used to align other things in the future? >> > For >> > > >> > example, >> > > >> > >> align read speed, I have >> > > >> > >> seen users requesting global rate limits. This feature may >> also >> > > need >> > > >> an >> > > >> > >> interface like this. >> > > >> > >> If we don't plan to extend this interface to support align >> other >> > > >> > things, I >> > > >> > >> suggest explicitly declaring >> > > >> > >> the purpose of the methods, such as `alignWatermarksForSplits` >> > > >> instead >> > > >> > of >> > > >> > >> `alignSplits`. >> > > >> > >> >> > > >> > >> This is a good point. Naming wise, it would usually be more >> > > >> extensible >> > > >> > to >> > > >> > >> just describe what the method actually does, instead of >> assuming >> > > the >> > > >> > >> purpose of doing this. For example, in this case, >> > > >> pauseOrResumeSplits() >> > > >> > >> would be more extensible because this can be used for any >> kind of >> > > >> flow >> > > >> > >> control, be it watermark alignment or simple rate limiting. >> > > >> > >> >> > > >> > >> 4. Interface or Method. >> > > >> > >> >> > > >> > >> I don't have a strong opinion on this. I think they have their >> > own >> > > >> > >> advantages. >> > > >> > >> In Flink SQL, we heavily use Interfaces for extending >> abilities >> > > >> > >> (SupportsXxxx) for TableSource/TableSink, >> > > >> > >> and I prefer Interfaces rather than methods in this case. When >> > you >> > > >> have >> > > >> > a >> > > >> > >> bunch of abilities and each ability >> > > >> > >> has more than one method, Interfaces can help to organize them >> > and >> > > >> make >> > > >> > >> users clear which methods >> > > >> > >> need to implement when you want to have an ability. >> > > >> > >> >> > > >> > >> I am OK with decorative interfaces if this is a general design >> > > >> pattern >> > > >> > in >> > > >> > >> the other components in Flink. But it looks like the current >> API >> > > >> > proposal >> > > >> > >> is not symmetric. >> > > >> > >> >> > > >> > >> The current proposal is essentially "MySourceReader implements >> > > >> > >> SourceReader, WithSplitsAlignment", along with "MySplitReader >> > > >> implements >> > > >> > >> AlignedSplitsReader". >> > > >> > >> >> > > >> > >> Should we make the API symmetric? I'd consider any one of the >> > > >> following >> > > >> > as >> > > >> > >> symmetric. >> > > >> > >> >> > > >> > >> a) "MySourceReader implements SourceReader, >> WithSplitsAlignment", >> > > >> along >> > > >> > >> with "MySplitReader implements SplitReader, >> WithSplitsAlignment", >> > > or >> > > >> > >> b) "MySourceReader implements AlignedSourceReader" and >> > > "MySplitReader >> > > >> > >> implements AlignedSplitReader", or >> > > >> > >> c) "MySourceReader implements SourceReader" and "MySplitReader >> > > >> > implements >> > > >> > >> SplitReader". >> > > >> > >> >> > > >> > >> I think the latest proposal according to Dawid would be: >> > > >> > >> d) "MySourceReader implements SourceReader" and "MySplitReader >> > > >> > implements >> > > >> > >> AlignedSplitReader". >> > > >> > >> I am fine with this API, although personally speaking I think >> it >> > is >> > > >> > simpler >> > > >> > >> to just add a new method to the split reader with default >> impl. >> > > >> > >> >> > > >> > >> @Dawid Wysakowicz <dwysakow...@apache.org> < >> > dwysakow...@apache.org >> > > > >> > > >> < >> > > >> > dwysakow...@apache.org> <dwysakow...@apache.org>, thanks for the >> > > reply. >> > > >> > >> >> > > >> > >> Having said that, as I don't have a preference and I agree >> most >> > of >> > > >> the >> > > >> > >> >> > > >> > >> sources will support the alignment I am fine following your >> > > >> suggestion >> > > >> > to >> > > >> > >> have the SourceReader extending from >> > WithWatermarksSplitsAlignment, >> > > >> but >> > > >> > >> would put the "supportsXXX" there, not in the Source to keep >> the >> > > two >> > > >> > >> methods together. >> > > >> > >> >> > > >> > >> One benefit of having the "supportsXXX" in Source is that this >> > > allows >> > > >> > some >> > > >> > >> compile time check. For example, if a user enabled watermark >> > > >> alignment >> > > >> > >> while it is not supported by the Source, an exception can be >> > thrown >> > > >> at >> > > >> > >> compile time. It seems in general useful. That said, I agree >> that >> > > API >> > > >> > >> cleanliness wise it is better to put the two methods together. >> > > >> > >> >> > > >> > >> Thanks, >> > > >> > >> >> > > >> > >> Jiangjie (Becket) Qin >> > > >> > >> >> > > >> > >> On Mon, Apr 25, 2022 at 5:56 PM Jark Wu <imj...@gmail.com> < >> > > >> > imj...@gmail.com> <imj...@gmail.com> <imj...@gmail.com> wrote: >> > > >> > >> >> > > >> > >> >> > > >> > >> Thank Dawid for the reminder on FLIP-182. Sorry I did miss it. >> > > >> > >> I don't have other concerns then. >> > > >> > >> >> > > >> > >> Best, >> > > >> > >> Jark >> > > >> > >> >> > > >> > >> On Mon, 25 Apr 2022 at 15:40, Dawid Wysakowicz < >> > > >> dwysakow...@apache.org> >> > > >> > <dwysakow...@apache.org> <dwysakow...@apache.org> < >> > > >> dwysakow...@apache.org> >> > > >> > >> wrote: >> > > >> > >> >> > > >> > >> >> > > >> > >> @Jark: >> > > >> > >> >> > > >> > >> 1. Will the framework always align with watermarks when the >> > source >> > > >> > >> implements the interface? >> > > >> > >> I'm afraid not every case needs watermark alignment even if >> Kafka >> > > >> > >> implements the interface, >> > > >> > >> and this will affect the throughput somehow. I agree with >> Becket >> > > >> > >> we may need a >> > > >> > >> `supportSplitsAlignment()` method for users to configure the >> > source >> > > >> to >> > > >> > >> enable/disable the alignment. >> > > >> > >> >> > > >> > >> 2. How does the framework calculate maxDesiredWatermark? >> > > >> > >> I think the algorithm of maxDesiredWatermark will greatly >> affect >> > > >> > >> >> > > >> > >> throughput >> > > >> > >> >> > > >> > >> if the reader is constantly >> > > >> > >> switching between pause and resume. Can users configure the >> > > >> alignment >> > > >> > >> offset? >> > > >> > >> >> > > >> > >> >> > > >> > >> This is covered in the previous FLIP[1] which has been already >> > > >> > >> >> > > >> > >> implemented >> > > >> > >> >> > > >> > >> in 1.15. In short, it must be enabled with the watermark >> strategy >> > > >> which >> > > >> > >> also configures drift and update interval. >> > > >> > >> >> > > >> > >> If we don't plan to extend this interface to support align >> other >> > > >> things, >> > > >> > >> >> > > >> > >> I >> > > >> > >> >> > > >> > >> suggest explicitly declaring >> > > >> > >> the purpose of the methods, such as `alignWatermarksForSplits` >> > > >> instead >> > > >> > of >> > > >> > >> `alignSplits`. >> > > >> > >> >> > > >> > >> >> > > >> > >> Sure let's rename it. >> > > >> > >> >> > > >> > >> @Becket: >> > > >> > >> >> > > >> > >> I understand your point. On the other hand putting all >> methods, >> > > even >> > > >> > with >> > > >> > >> "supportsXXX" methods for enabling certain features, makes the >> > > entry >> > > >> > >> threshold for writing a new source higher. Instead of >> focusing on >> > > the >> > > >> > >> >> > > >> > >> basic >> > > >> > >> >> > > >> > >> and required properties of the Source, the person >> implementing a >> > > >> source >> > > >> > >> must bother with and need to figure out what all of the extra >> > > >> features >> > > >> > >> >> > > >> > >> are >> > > >> > >> >> > > >> > >> about and how to deal with them. It makes it also harder to >> > > organize >> > > >> > >> methods in coupled groups as Jark said. >> > > >> > >> >> > > >> > >> Having said that, as I don't have a preference and I agree >> most >> > of >> > > >> the >> > > >> > >> sources will support the alignment I am fine following your >> > > >> suggestion >> > > >> > to >> > > >> > >> have the SourceReader extending from >> > WithWatermarksSplitsAlignment, >> > > >> but >> > > >> > >> would put the "supportsXXX" there, not in the Source to keep >> the >> > > two >> > > >> > >> methods together. >> > > >> > >> >> > > >> > >> Lastly, I agree it is really unfortunate the "alignSplits" >> > methods >> > > >> > differ >> > > >> > >> slightly for SourceReader and SpitReader. The reason for that >> is >> > > >> > >> SourceReaderBase deals only with SplitIds, whereas SplitReader >> > > needs >> > > >> the >> > > >> > >> actual splits to pause them. I found the discrepancy >> acceptable >> > for >> > > >> the >> > > >> > >> sake of simplifying changes significantly, especially as they >> > would >> > > >> > >> >> > > >> > >> highly >> > > >> > >> >> > > >> > >> likely impact performance as we would have to perform >> additional >> > > >> > lookups. >> > > >> > >> Moreover the SplitReader is a secondary interface. >> > > >> > >> >> > > >> > >> Best, >> > > >> > >> >> > > >> > >> Dawid >> > > >> > >> >> > > >> > >> [1] https://cwiki.apache.org/confluence/x/hQYBCw >> > > >> > >> >> > > >> > >> On 24/04/2022 17:15, Jark Wu wrote: >> > > >> > >> >> > > >> > >> Thanks for the effort, Dawid and Sebastian! >> > > >> > >> >> > > >> > >> I just have some minor questions (maybe I missed something). >> > > >> > >> >> > > >> > >> 1. Will the framework always align with watermarks when the >> > source >> > > >> > >> implements the interface? >> > > >> > >> I'm afraid not every case needs watermark alignment even if >> Kafka >> > > >> > >> implements the interface, >> > > >> > >> and this will affect the throughput somehow. I agree with >> Becket >> > > >> > >> we may need a >> > > >> > >> `supportSplitsAlignment()` method for users to configure the >> > source >> > > >> to >> > > >> > >> enable/disable the alignment. >> > > >> > >> >> > > >> > >> 2. How does the framework calculate maxDesiredWatermark? >> > > >> > >> I think the algorithm of maxDesiredWatermark will greatly >> affect >> > > >> > >> >> > > >> > >> throughput >> > > >> > >> >> > > >> > >> if the reader is constantly >> > > >> > >> switching between pause and resume. Can users configure the >> > > >> alignment >> > > >> > >> offset? >> > > >> > >> >> > > >> > >> 3. Interface/Method Name. >> > > >> > >> Can the interface be used to align other things in the future? >> > For >> > > >> > >> >> > > >> > >> example, >> > > >> > >> >> > > >> > >> align read speed, I have >> > > >> > >> seen users requesting global rate limits. This feature may >> also >> > > need >> > > >> an >> > > >> > >> interface like this. >> > > >> > >> If we don't plan to extend this interface to support align >> other >> > > >> things, >> > > >> > >> >> > > >> > >> I >> > > >> > >> >> > > >> > >> suggest explicitly declaring >> > > >> > >> the purpose of the methods, such as `alignWatermarksForSplits` >> > > >> instead >> > > >> > of >> > > >> > >> `alignSplits`. >> > > >> > >> >> > > >> > >> 4. Interface or Method. >> > > >> > >> I don't have a strong opinion on this. I think they have their >> > own >> > > >> > >> advantages. >> > > >> > >> In Flink SQL, we heavily use Interfaces for extending >> abilities >> > > >> > >> (SupportsXxxx) for TableSource/TableSink, >> > > >> > >> and I prefer Interfaces rather than methods in this case. When >> > you >> > > >> have >> > > >> > a >> > > >> > >> bunch of abilities and each ability >> > > >> > >> has more than one method, Interfaces can help to organize them >> > and >> > > >> make >> > > >> > >> users clear which methods >> > > >> > >> need to implement when you want to have an ability. >> > > >> > >> >> > > >> > >> >> > > >> > >> Best, >> > > >> > >> Jark >> > > >> > >> >> > > >> > >> On Sun, 24 Apr 2022 at 18:13, Becket Qin < >> becket....@gmail.com> >> > < >> > > >> > becket....@gmail.com> <becket....@gmail.com> < >> becket....@gmail.com> >> > < >> > > >> > >> becket....@gmail.com> wrote: >> > > >> > >> >> > > >> > >> Hi Dawid, >> > > >> > >> >> > > >> > >> Thanks for the explanation. Apologies that I somehow misread a >> > > bunch >> > > >> of >> > > >> > >> "align" and thought they were "assign". >> > > >> > >> >> > > >> > >> Regarding 1, by default implementation, I was thinking of the >> > > default >> > > >> > >> >> > > >> > >> no-op >> > > >> > >> >> > > >> > >> implementation. I am a little worried about the proliferation >> of >> > > >> > >> >> > > >> > >> decorative >> > > >> > >> >> > > >> > >> interfaces. I think the most important thing about interfaces >> is >> > > that >> > > >> > >> >> > > >> > >> they >> > > >> > >> >> > > >> > >> are easy to understand. In this case, I prefer adding new >> method >> > to >> > > >> the >> > > >> > >> existing interface for the following reasons: >> > > >> > >> >> > > >> > >> a) I feel the biggest drawback of decorative interfaces is >> which >> > > >> > >> >> > > >> > >> interface >> > > >> > >> >> > > >> > >> they can decorate and which combinations of multiple >> decorative >> > > >> > >> >> > > >> > >> interfaces >> > > >> > >> >> > > >> > >> are valid. In the current FLIP, the withSplitsAlignment >> interface >> > > is >> > > >> > only >> > > >> > >> applicable to the SourceReader which means it can't decorate >> any >> > > >> other >> > > >> > >> interface. From an interface design perspective, a natural >> > question >> > > >> is >> > > >> > >> >> > > >> > >> why >> > > >> > >> >> > > >> > >> not let "AlignedSplitReader" extend "withSplitsAlignment"? >> And it >> > > is >> > > >> > also >> > > >> > >> natural to assume that a split reader implementing both >> > SplitReader >> > > >> and >> > > >> > >> WithSplitAlignment would work, because a source reader >> > implementing >> > > >> > >> SourceReader and withSplitsAlignment works. So why isn't >> there an >> > > >> > >> >> > > >> > >> interface >> > > >> > >> >> > > >> > >> of AlignedSourceReader? In the future, if there is a new >> feature >> > > >> added >> > > >> > >> (e.g. sorted or pre-partitioned data aware), are we going to >> > create >> > > >> > >> >> > > >> > >> another >> > > >> > >> >> > > >> > >> interface of SplitReader such as SortedSplitReader or >> > > >> > >> >> > > >> > >> PrePartitionedAware? >> > > >> > >> >> > > >> > >> Can they be combined? So I think the additional decorative >> > > interface >> > > >> > like >> > > >> > >> withSplitsAlignment actually increases the understanding cost >> of >> > > >> users >> > > >> > >> because they have to know what decorative interfaces are >> there, >> > > which >> > > >> > >> interface they can decorate and which combinations of the >> > > decorative >> > > >> > >> interfaces are valid and which are not. Ideally we want to >> avoid >> > > >> that. >> > > >> > To >> > > >> > >> be clear, I am not opposing having an interface of >> > > >> withSplitsAlignment, >> > > >> > >> >> > > >> > >> it >> > > >> > >> >> > > >> > >> is completely OK to have it as an internal interface and let >> > > >> > SourceReader >> > > >> > >> and SplitReader both extend it. >> > > >> > >> >> > > >> > >> b) Adding a new method to the SourceReader with a default >> > > >> implementation >> > > >> > >> >> > > >> > >> of >> > > >> > >> >> > > >> > >> no-op would help avoid logic branching in the source logic, >> > > >> especially >> > > >> > >> given that we agree that the vast majority of the SourceReader >> > > >> > >> implementations, if not all, would just extend from the >> > > >> > SourceReaderBase. >> > > >> > >> That means adding a new method to the interface would >> effectively >> > > >> give >> > > >> > >> >> > > >> > >> the >> > > >> > >> >> > > >> > >> same user experience, but simpler. >> > > >> > >> >> > > >> > >> c) A related design principle that may be worth discussing is >> how >> > > do >> > > >> we >> > > >> > >> >> > > >> > >> let >> > > >> > >> >> > > >> > >> the Source implementations tell Flink what capability is >> > supported >> > > >> and >> > > >> > >> >> > > >> > >> what >> > > >> > >> >> > > >> > >> is not. Personally speaking I feel the most intuitive place >> to me >> > > is >> > > >> in >> > > >> > >> >> > > >> > >> the >> > > >> > >> >> > > >> > >> Source itself, because that is the entrance of the entire >> Source >> > > >> > >> >> > > >> > >> connector >> > > >> > >> >> > > >> > >> logic. >> > > >> > >> >> > > >> > >> Based on the above thoughts, I am wondering if the following >> > > >> interface >> > > >> > >> would be easier to understand by the users. >> > > >> > >> >> > > >> > >> - Change "withSplitsAlignment" to internal interface, let both >> > > >> > >> >> > > >> > >> SourceReader >> > > >> > >> >> > > >> > >> and SplitReader extend from it, with a default no-op >> > > implementation. >> > > >> > >> - Add a new method "boolean supportSplitsAlignment()" to the >> > Source >> > > >> > >> interface, with a default implementation returning false. >> Sources >> > > >> that >> > > >> > >> >> > > >> > >> have >> > > >> > >> >> > > >> > >> implemented the alignment logic can change this to return >> true, >> > and >> > > >> > >> override the alignSplits() methods in the SourceReader / >> > > SplitReader >> > > >> if >> > > >> > >> needed. >> > > >> > >> - In the future, if a new optional feature is going to be >> added >> > to >> > > >> the >> > > >> > >> Source, and that feature requires the awareness from Flink, we >> > can >> > > >> add >> > > >> > >> >> > > >> > >> more >> > > >> > >> >> > > >> > >> such methods to the Source. >> > > >> > >> >> > > >> > >> What do you think? >> > > >> > >> >> > > >> > >> Thanks, >> > > >> > >> >> > > >> > >> Jiangjie (Becket) Qin >> > > >> > >> >> > > >> > >> >> > > >> > >> >> > > >> > >> >> > > >> > >> >> > > >> > >> On Fri, Apr 22, 2022 at 4:05 PM Dawid Wysakowicz < >> > > >> > dwysakow...@apache.org> <dwysakow...@apache.org> < >> > > >> dwysakow...@apache.org> >> > > >> > <dwysakow...@apache.org> >> > > >> > >> <dwysakow...@apache.org> <dwysakow...@apache.org> < >> > > >> > dwysakow...@apache.org> <dwysakow...@apache.org> >> > > >> > >> >> > > >> > >> wrote: >> > > >> > >> >> > > >> > >> >> > > >> > >> @Konstantin: >> > > >> > >> >> > > >> > >> As part of this FLIP, the `AlignedSplitReader` interface (aka >> the >> > > >> stop & >> > > >> > >> resume behavior) will be implemented for Kafka and Pulsar >> only, >> > > >> correct? >> > > >> > >> >> > > >> > >> Correct, as far as I know though, those are the only sources >> > which >> > > >> > >> >> > > >> > >> consume >> > > >> > >> >> > > >> > >> concurrently from multiple splits and thus alignment applies. >> > > >> > >> >> > > >> > >> @Thomas: >> > > >> > >> >> > > >> > >> I wonder if "supporting" split alignment in SourceReaderBase >> and >> > > then >> > > >> > >> >> > > >> > >> doing >> > > >> > >> >> > > >> > >> nothing if the split reader does not implement >> AlignedSplitReader >> > > >> could >> > > >> > >> >> > > >> > >> be >> > > >> > >> >> > > >> > >> misleading? Perhaps WithSplitsAlignment can instead be added >> to >> > the >> > > >> > >> specific source reader (i.e. KafkaSourceReader) to make it >> > explicit >> > > >> that >> > > >> > >> the source actually supports it. >> > > >> > >> >> > > >> > >> I understand your concern. Hmm, I think we could actually do >> > that. >> > > >> Given >> > > >> > >> the actual implementation of the SourceReaderBase#alignSplits >> is >> > > >> rather >> > > >> > >> short (just a forward to the corresponding method of >> > SplitFetcher), >> > > >> we >> > > >> > >> could reimplement it in the actual source implementations. >> This >> > > >> solution >> > > >> > >> has the downside though. Authors of new sources would have to >> do >> > > two >> > > >> > >> things: extend from AlignedSplitReader and implement >> > > >> > >> >> > > >> > >> WithSplitsAssignment, >> > > >> > >> >> > > >> > >> instead of just extending AlignedSplitReader. I would be fine >> > with >> > > >> such >> > > >> > a >> > > >> > >> tradeoff though. What others think? >> > > >> > >> >> > > >> > >> @Steven: >> > > >> > >> >> > > >> > >> For this part from the motivation section, is it accurate? >> Let's >> > > >> assume >> > > >> > >> >> > > >> > >> one >> > > >> > >> >> > > >> > >> source task consumes from 3 partitions and one of the >> partition >> > is >> > > >> > >> significantly slower. In this situation, watermark for this >> > source >> > > >> task >> > > >> > >> won't hold back as it is reading recent data from other two >> Kafka >> > > >> > >> partitions. As a result, it won't hold back the overall >> > watermark. >> > > I >> > > >> > >> thought the problem is that we may have late data for this >> slow >> > > >> > >> >> > > >> > >> partition. >> > > >> > >> >> > > >> > >> It will hold back the watermark. Watermark of an operator is >> the >> > > >> minimum >> > > >> > >> of watermarks of all splits[1] >> > > >> > >> >> > > >> > >> I have another question about the restart. Say split >> alignment is >> > > >> > >> triggered. checkpoint is completed. job failed and restored >> from >> > > the >> > > >> > last >> > > >> > >> checkpoint. because alignment decision is not checkpointed, >> > > initially >> > > >> > >> alignment won't be enforced until we get a cycle of watermark >> > > >> > aggregation >> > > >> > >> and propagation, right? Not saying this corner is a problem. >> Just >> > > >> want >> > > >> > to >> > > >> > >> understand it more. >> > > >> > >> >> > > >> > >> Your understanding is correct. >> > > >> > >> >> > > >> > >> @Becket: >> > > >> > >> >> > > >> > >> 1. I think watermark alignment is sort of a general use case, >> so >> > > >> should >> > > >> > >> >> > > >> > >> we >> > > >> > >> >> > > >> > >> just add the related methods to SourceReader directly instead >> of >> > > >> > >> introducing the new interface of WithSplitAssignment? We can >> > > provide >> > > >> > >> default implementations, so backwards compatibility won't be >> an >> > > >> issue. >> > > >> > >> >> > > >> > >> I don't think we can provide a default implementation. How >> would >> > we >> > > >> do >> > > >> > >> that? Would it be just a no-op? Is it better than having an >> > opt-in >> > > >> > >> interface? The default implementation would have to be added >> > > >> exclusively >> > > >> > >> >> > > >> > >> in >> > > >> > >> >> > > >> > >> a *Public* SourceReader interface. By the way notice >> > > SourceReaderBase >> > > >> > >> does extend from WithSplitsAlignment, so effectively all >> > > >> implementations >> > > >> > >> >> > > >> > >> do >> > > >> > >> >> > > >> > >> handle the alignment case. To be honest I think it is >> impossible >> > to >> > > >> > >> implement the SourceReader interface directly by end users. >> > > >> > >> >> > > >> > >> 2. As you mentioned, the SplitReader interface probably also >> > needs >> > > >> some >> > > >> > >> change to support throttling at the split granularity. Can you >> > add >> > > >> that >> > > >> > >> interface change into the public interface section as well? >> > > >> > >> >> > > >> > >> It has been added from the beginning. See >> *AlignedSplitReader.* >> > > >> > >> >> > > >> > >> 3. Nit, can we avoid using the method name assignSplits here, >> > given >> > > >> that >> > > >> > >> >> > > >> > >> it >> > > >> > >> >> > > >> > >> is not actually changing the split assignments? It seems >> > something >> > > >> like >> > > >> > >> pauseOrResumeSplits(), or adjustSplitsThrottling() is more >> > > accurate. >> > > >> > >> >> > > >> > >> The method's called *alignSplits*, not assign. Do you still >> > prefer >> > > a >> > > >> > >> different name for that? Personally, I am open for suggestions >> > > here. >> > > >> > >> >> > > >> > >> Best, >> > > >> > >> >> > > >> > >> Dawid >> > > >> > >> >> > > >> > >> [1] >> > > >> > >> >> > > >> > >> >> > > >> > >> >> > > >> > >> >> > > >> > >> > > >> >> > > >> > >> https://nightlies.apache.org/flink/flink-docs-master/docs/dev/datastream/sources/#watermark-generation >> > > >> > >> >> > > >> > >> On 22/04/2022 05:59, Becket Qin wrote: >> > > >> > >> >> > > >> > >> Thanks for driving the effort, Sebastion. I think the >> motivation >> > > >> makes a >> > > >> > >> lot of sense. Just a few suggestions / questions. >> > > >> > >> >> > > >> > >> 1. I think watermark alignment is sort of a general use case, >> so >> > > >> should >> > > >> > >> >> > > >> > >> we >> > > >> > >> >> > > >> > >> just add the related methods to SourceReader directly instead >> of >> > > >> > >> introducing the new interface of WithSplitAssignment? We can >> > > provide >> > > >> > >> default implementations, so backwards compatibility won't be >> an >> > > >> issue. >> > > >> > >> >> > > >> > >> 2. As you mentioned, the SplitReader interface probably also >> > needs >> > > >> some >> > > >> > >> change to support throttling at the split granularity. Can you >> > add >> > > >> that >> > > >> > >> interface change into the public interface section as well? >> > > >> > >> >> > > >> > >> 3. Nit, can we avoid using the method name assignSplits here, >> > given >> > > >> that >> > > >> > >> >> > > >> > >> it >> > > >> > >> >> > > >> > >> is not actually changing the split assignments? It seems >> > something >> > > >> like >> > > >> > >> pauseOrResumeSplits(), or adjustSplitsThrottling() is more >> > > accurate. >> > > >> > >> >> > > >> > >> Thanks, >> > > >> > >> >> > > >> > >> Jiangjie (Becket) Qin >> > > >> > >> >> > > >> > >> On Thu, Apr 21, 2022 at 11:39 PM Steven Wu < >> stevenz...@gmail.com >> > > >> > > < >> > > >> > stevenz...@gmail.com> <stevenz...@gmail.com> < >> stevenz...@gmail.com> >> > < >> > > >> > >> stevenz...@gmail.com> < >> > > >> > >> stevenz...@gmail.com> wrote: >> > > >> > >> >> > > >> > >> However, a single source operator may read data from multiple >> > > >> > >> >> > > >> > >> splits/partitions, e.g., multiple Kafka partitions, such that >> > even >> > > >> with >> > > >> > >> watermark alignment the source operator may need to buffer >> > > excessive >> > > >> > >> >> > > >> > >> amount >> > > >> > >> >> > > >> > >> of data if one split emits data faster than another. >> > > >> > >> >> > > >> > >> For this part from the motivation section, is it accurate? >> Let's >> > > >> assume >> > > >> > >> >> > > >> > >> one >> > > >> > >> >> > > >> > >> source task consumes from 3 partitions and one of the >> partition >> > is >> > > >> > >> significantly slower. In this situation, watermark for this >> > source >> > > >> task >> > > >> > >> won't hold back as it is reading recent data from other two >> Kafka >> > > >> > >> partitions. As a result, it won't hold back the overall >> > watermark. >> > > I >> > > >> > >> thought the problem is that we may have late data for this >> slow >> > > >> > >> >> > > >> > >> partition. >> > > >> > >> >> > > >> > >> I have another question about the restart. Say split >> alignment is >> > > >> > >> triggered. checkpoint is completed. job failed and restored >> from >> > > the >> > > >> > last >> > > >> > >> checkpoint. because alignment decision is not checkpointed, >> > > initially >> > > >> > >> alignment won't be enforced until we get a cycle of watermark >> > > >> > aggregation >> > > >> > >> and propagation, right? Not saying this corner is a problem. >> Just >> > > >> want >> > > >> > to >> > > >> > >> understand it more. >> > > >> > >> >> > > >> > >> >> > > >> > >> >> > > >> > >> On Thu, Apr 21, 2022 at 8:20 AM Thomas Weise <t...@apache.org> >> < >> > > >> > t...@apache.org> <t...@apache.org> <t...@apache.org> < >> > > >> > >> t...@apache.org> < >> > > >> > >> t...@apache.org> wrote: >> > > >> > >> >> > > >> > >> Thanks for working on this! >> > > >> > >> >> > > >> > >> I wonder if "supporting" split alignment in SourceReaderBase >> and >> > > then >> > > >> > >> >> > > >> > >> doing >> > > >> > >> >> > > >> > >> nothing if the split reader does not implement >> AlignedSplitReader >> > > >> could >> > > >> > >> >> > > >> > >> be >> > > >> > >> >> > > >> > >> misleading? Perhaps WithSplitsAlignment can instead be added >> to >> > the >> > > >> > >> specific source reader (i.e. KafkaSourceReader) to make it >> > explicit >> > > >> that >> > > >> > >> the source actually supports it. >> > > >> > >> >> > > >> > >> Thanks, >> > > >> > >> Thomas >> > > >> > >> >> > > >> > >> >> > > >> > >> On Thu, Apr 21, 2022 at 4:57 AM Konstantin Knauf < >> > > kna...@apache.org> >> > > >> < >> > > >> > kna...@apache.org> <kna...@apache.org> <kna...@apache.org> < >> > > >> > >> kna...@apache.org> < >> > > >> > >> kna...@apache.org> >> > > >> > >> >> > > >> > >> wrote: >> > > >> > >> >> > > >> > >> >> > > >> > >> Hi Sebastian, Hi Dawid, >> > > >> > >> >> > > >> > >> As part of this FLIP, the `AlignedSplitReader` interface (aka >> the >> > > >> stop >> > > >> > >> >> > > >> > >> & >> > > >> > >> >> > > >> > >> resume behavior) will be implemented for Kafka and Pulsar >> only, >> > > >> > >> >> > > >> > >> correct? >> > > >> > >> >> > > >> > >> +1 in general. I believe it is valuable to complete the >> watermark >> > > >> > >> >> > > >> > >> aligned >> > > >> > >> >> > > >> > >> story with this FLIP. >> > > >> > >> >> > > >> > >> Cheers, >> > > >> > >> >> > > >> > >> Konstantin >> > > >> > >> >> > > >> > >> >> > > >> > >> >> > > >> > >> >> > > >> > >> >> > > >> > >> >> > > >> > >> >> > > >> > >> On Thu, Apr 21, 2022 at 12:36 PM Dawid Wysakowicz < >> > > >> > dwysakow...@apache.org> <dwysakow...@apache.org> >> > > >> > >> >> > > >> > >> wrote: >> > > >> > >> >> > > >> > >> >> > > >> > >> To be explicit, having worked on it, I support it ;) I think >> we >> > can >> > > >> > >> start a vote thread soonish, as there are no concerns so far. >> > > >> > >> >> > > >> > >> Best, >> > > >> > >> >> > > >> > >> Dawid >> > > >> > >> >> > > >> > >> On 13/04/2022 11:27, Sebastian Mattheis wrote: >> > > >> > >> >> > > >> > >> Dear Flink developers, >> > > >> > >> >> > > >> > >> I would like to open a discussion on FLIP 217 [1] for an >> > extension >> > > >> > >> >> > > >> > >> of >> > > >> > >> >> > > >> > >> Watermark Alignment to perform alignment also in >> SplitReaders. To >> > > >> > >> >> > > >> > >> do >> > > >> > >> >> > > >> > >> so, >> > > >> > >> >> > > >> > >> SplitReaders must be able to suspend and resume reading from >> > split >> > > >> > >> >> > > >> > >> sources >> > > >> > >> >> > > >> > >> where the SourceOperator coordinates and controlls suspend and >> > > >> > >> >> > > >> > >> resume. >> > > >> > >> >> > > >> > >> To >> > > >> > >> >> > > >> > >> gather information about current watermarks of the >> SplitReaders, >> > we >> > > >> > >> >> > > >> > >> extend >> > > >> > >> >> > > >> > >> the internal WatermarkOutputMulitplexer and report watermarks >> to >> > > >> > >> >> > > >> > >> the >> > > >> > >> >> > > >> > >> SourceOperator. >> > > >> > >> >> > > >> > >> There is a PoC for this FLIP [2], prototyped by Arvid Heise >> and >> > > >> > >> >> > > >> > >> revised >> > > >> > >> >> > > >> > >> and >> > > >> > >> >> > > >> > >> reworked by Dawid Wysakowicz (He did most of the work.) and >> me. >> > The >> > > >> > >> >> > > >> > >> changes >> > > >> > >> >> > > >> > >> are backwards compatible in a way that if affected components >> do >> > > >> > >> >> > > >> > >> not >> > > >> > >> >> > > >> > >> support split alignment the behavior is as before. >> > > >> > >> >> > > >> > >> Best, >> > > >> > >> Sebastian >> > > >> > >> >> > > >> > >> [1] >> > > >> > >> >> > > >> > >> >> > > >> > >> >> > > >> > >> >> > > >> > >> >> > > >> > >> >> > > >> > >> > > >> >> > > >> > >> https://cwiki.apache.org/confluence/display/FLINK/FLIP-217+Support+watermark+alignment+of+source+splits >> > > >> > >> >> > > >> > >> [2] https://github.com/dawidwys/flink/tree/aligned-splits >> > > >> > >> >> > > >> > >> -- >> > > >> > >> >> > > >> > >> Konstantin Knaufhttps:// >> > > >> twitter.com/snntrablehttps://github.com/knaufk >> > > >> > >> >> > > >> > >> >> > > >> > >> > > >> >> > > > >> > > >> > >> >
